CN116824314A - Information acquisition methods and systems - Google Patents

Information acquisition methods and systems Download PDF

Info

Publication number
CN116824314A
CN116824314A CN202210272796.0A CN202210272796A CN116824314A CN 116824314 A CN116824314 A CN 116824314A CN 202210272796 A CN202210272796 A CN 202210272796A CN 116824314 A CN116824314 A CN 116824314A
Authority
CN
China
Prior art keywords
information
verification
recognition model
user
image recognition
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210272796.0A
Other languages
Chinese (zh)
Inventor
胡颉
陈宇阳
俞新华
刘智
张焓
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Mobile Communications Group Co Ltd
China Mobile Group Jiangsu Co Ltd
Original Assignee
China Mobile Communications Group Co Ltd
China Mobile Group Jiangsu Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by China Mobile Communications Group Co Ltd, China Mobile Group Jiangsu Co Ltd filed Critical China Mobile Communications Group Co Ltd
Priority to CN202210272796.0A priority Critical patent/CN116824314A/en
Publication of CN116824314A publication Critical patent/CN116824314A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/08Learning methods
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/08Learning methods
    • G06N3/084Backpropagation, e.g. using gradient descent

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Molecular Biology (AREA)
  • Artificial Intelligence (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Image Analysis (AREA)
  • Collating Specific Patterns (AREA)

Abstract

The invention provides an information acquisition method and system, which can be applied to the field of payment security, wherein the method comprises the following steps: pre-training the convolutional neural network by using a first training sample to obtain an image recognition model; performing optimization training on the image recognition model by using a second training sample to obtain a target image recognition model; carrying out face verification on the received identity verification information uploaded by the user terminal, and carrying out position verification on the identity verification information based on a target image recognition model; and under the condition that the face verification and the position verification are both passed, transmitting the received data information corresponding to the information acquisition request transmitted by the user terminal to the user terminal. The system is used for executing the method. The invention can realize double security guarantee of user identity and user environment through double identity authentication of face authentication and position authentication, and avoid property loss of users caused by authentication of user identity based on image processing means of some illegal user terminals.

Description

信息获取方法及系统Information acquisition methods and systems

技术领域Technical field

本发明涉及信息处理技术领域,尤其涉及一种信息获取方法及系统。The present invention relates to the field of information processing technology, and in particular to an information acquisition method and system.

背景技术Background technique

现有技术中,人脸识别方式主要是单一的获取用户的脸部特征,然后将脸部特征与特征库中的特征进行匹配以实现身份识别,这种方式只能判断用户身份,并不能识别用户环境信息的安全,为不法用户终端基于图像处理手段通过用户身份验证,从而造成用户的财产损失。可见,如何消除消费或支付环境中存在的安全问题已经成为亟需解决的技术问题。In the existing technology, the face recognition method mainly obtains the user's facial features alone, and then matches the facial features with the features in the feature library to achieve identity recognition. This method can only determine the user's identity, but cannot identify the user. The security of user environment information requires illegal user terminals to pass user identity verification based on image processing methods, thereby causing user property losses. It can be seen that how to eliminate security issues in the consumption or payment environment has become an urgent technical issue that needs to be solved.

发明内容Contents of the invention

本发明提供的信息获取方法及系统,用于解决现有技术中存在的问题,通过人脸验证和位置验证的双重身份验证,可以实现用户身份和用户环境的双重安全保障,避免了一些不法用户终端,基于图像处理手段,通过用户身份的验证而造成用户的财产损失。The information acquisition method and system provided by the present invention are used to solve the problems existing in the existing technology. Through dual identity verification of face verification and location verification, dual security guarantees of user identity and user environment can be achieved, and some unscrupulous users can be avoided. The terminal, based on image processing means, causes the user's property loss through user identity verification.

本发明提供的一种信息获取方法,包括:An information acquisition method provided by the invention includes:

使用第一训练样本对卷积神经网络进行预训练,以获取图像识别模型;Pre-train the convolutional neural network using the first training sample to obtain the image recognition model;

使用第二训练样本对所述图像识别模型进行优化训练,以获取目标图像识别模型;Use the second training sample to optimize and train the image recognition model to obtain the target image recognition model;

对接收到的用户终端上传的身份验证信息进行人脸验证以及基于所述目标图像识别模型对所述身份验证信息进行位置验证;Perform face verification on the received identity verification information uploaded by the user terminal and perform location verification on the identity verification information based on the target image recognition model;

在人脸验证和位置验证均通过的情况下,将接收到的所述用户终端发送的信息获取请求对应的数据信息发送给所述用户终端;If both the face verification and the location verification pass, send the data information corresponding to the received information acquisition request sent by the user terminal to the user terminal;

其中,所述第一训练样本是根据获取的多数的环境图像中每一环境图像对应的第一位置信息对每一环境图像进行标注后得到的;Wherein, the first training sample is obtained by labeling each environment image according to the first position information corresponding to each environment image in the plurality of acquired environment images;

所述第二训练样本是根据人脸图像与所述第一训练样本中的环境图像叠加后的目标图像确定的;The second training sample is determined based on the target image obtained by superimposing the face image and the environment image in the first training sample;

所述身份验证信息包括用户实时拍摄的包括用户人脸图像的环境图像和用户的位置信息。The identity verification information includes an environment image including the user's face image taken by the user in real time and the user's location information.

根据本发明提供的一种信息获取方法,所述使用第一训练样本对卷积神经网络进行预训练,以获取图像识别模型,包括:According to an information acquisition method provided by the present invention, pre-training a convolutional neural network using a first training sample to obtain an image recognition model includes:

将所述第一训练样本输入至所述卷积神经网络进行预训练,并根据所述卷积神经网络输出的每一环境图像的第二位置信息与所述第一位置信息的比较结果,对所述卷积神经网络的超参数进行调整;The first training sample is input to the convolutional neural network for pre-training, and based on the comparison result between the second position information of each environmental image output by the convolutional neural network and the first position information, The hyperparameters of the convolutional neural network are adjusted;

根据调整后的卷积神经网络,确定所述图像识别模型。According to the adjusted convolutional neural network, the image recognition model is determined.

根据本发明提供的一种信息获取方法,所述对所述使用第二训练样本对所述图像识别模型进行优化训练,以获取目标图像识别模型,包括:According to an information acquisition method provided by the present invention, the optimization training of the image recognition model using the second training sample to obtain the target image recognition model includes:

将所述第二训练样本输入至所述图像识别模型进行优化训练,并根据所述图像识别模型输出的所述目标图像的第三位置信息与所述第一位置信息的比较结果,采用反向传播算法和随机梯度算法对所述图像识别模型进行优化,以获取所述目标图像识别模型。The second training sample is input to the image recognition model for optimization training, and according to the comparison result of the third position information of the target image output by the image recognition model and the first position information, a reverse The propagation algorithm and the stochastic gradient algorithm optimize the image recognition model to obtain the target image recognition model.

根据本发明提供的一种信息获取方法,所述用户终端上传的身份验证信息进行位置是通过如下方式确定的:According to an information acquisition method provided by the present invention, the location of the identity verification information uploaded by the user terminal is determined in the following manner:

在接收到所述用户终端发送的所述信息获取请求后,向所述用户终端发送身份验证信息采集信息,以使所述用户终端根据接收到的所述身份验证信息采集信息,采集所述身份验证信息。After receiving the information acquisition request sent by the user terminal, sending identity verification information collection information to the user terminal, so that the user terminal collects information according to the received identity verification information, and collects the identity verify message.

根据本发明提供的一种信息获取方法,所述对接收到的用户终端上传的身份验证信息进行人脸验证以及基于所述目标图像识别模型对所述身份验证信息进行位置验证,包括:According to an information acquisition method provided by the present invention, the face verification of the received identity verification information uploaded by the user terminal and the location verification of the identity verification information based on the target image recognition model include:

提取所述包括用户人脸图像的环境图像的第一人脸特征;Extracting the first facial feature of the environment image including the user's facial image;

将所述第一人脸特征与所述身份验证信息中用户注册时的第二人脸特征进行特征向量计算,以获取所述第一人脸特征与所述第二人脸特征之间的距离;Perform feature vector calculation on the first facial feature and the second facial feature when the user registered in the identity verification information to obtain the distance between the first facial feature and the second facial feature ;

根据所述距离,完成对所述身份验证信息的人脸验证;According to the distance, complete face verification of the identity verification information;

在人脸验证通过的情况下,基于所述目标图像识别模型对所述身份验证信息进行位置验证。If the face verification passes, the identity verification information is location verified based on the target image recognition model.

根据本发明提供的一种信息获取方法,所述基于所述目标图像识别模型对所述身份验证信息进行位置验证,包括:According to an information acquisition method provided by the present invention, the location verification of the identity verification information based on the target image recognition model includes:

将所述包括用户人脸图像的环境图像输入至所述目标图像识别模型,并根据所述目标图像识别模型输出的第四位置信息与所述用户的位置信息的比较结果,完成对所述身份验证信息的位置验证。The environment image including the user's face image is input to the target image recognition model, and based on the comparison result between the fourth position information output by the target image recognition model and the user's position information, the identity is completed. Location verification of verification information.

本发明还提供一种信息获取系统,包括:第一训练模块、第二训练模块、身份验证模块和信息获取模块;The invention also provides an information acquisition system, including: a first training module, a second training module, an identity verification module and an information acquisition module;

所述第一训练模块,用于使用第一训练样本对卷积神经网络进行预训练,以获取图像识别模型;The first training module is used to pre-train the convolutional neural network using the first training sample to obtain the image recognition model;

所述第二训练模块,用于使用第二训练样本对所述图像识别模型进行优化训练,以获取目标图像识别模型;The second training module is used to optimize and train the image recognition model using the second training sample to obtain the target image recognition model;

所述身份验证模块,用于对接收到的用户终端上传的身份验证信息进行人脸验证以及基于所述目标图像识别模型对所述身份验证信息进行位置验证;The identity verification module is used to perform face verification on the received identity verification information uploaded by the user terminal and perform location verification on the identity verification information based on the target image recognition model;

所述信息获取模块,用于在人脸验证和位置验证均通过的情况下,将接收到的所述用户终端发送的信息获取请求对应的数据信息发送给所述用户终端;The information acquisition module is configured to send data information corresponding to the received information acquisition request sent by the user terminal to the user terminal when both face verification and location verification pass;

其中,所述第一训练样本是根据获取的多数的环境图像中每一环境图像对应的第一位置信息对每一环境图像进行标注后得到的;Wherein, the first training sample is obtained by labeling each environment image according to the first position information corresponding to each environment image in the plurality of acquired environment images;

所述第二训练样本是根据人脸图像与所述第一训练样本中的环境图像叠加后的目标图像确定的;The second training sample is determined based on the target image obtained by superimposing the face image and the environment image in the first training sample;

所述身份验证信息包括用户实时拍摄的包括用户人脸图像的环境图像和用户的位置信息。The identity verification information includes an environment image including the user's face image taken by the user in real time and the user's location information.

本发明还提供一种电子设备,包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序,所述处理器执行所述程序时实现如上述任一种所述信息获取方法。The present invention also provides an electronic device, including a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements any of the above information acquisition methods. .

本发明还提供一种处理器可读存储介质,所述处理器可读存储介质存储有计算机程序,所述计算机程序用于使所述处理器执行,如上述任一种所述信息获取方法。The present invention also provides a processor-readable storage medium. The processor-readable storage medium stores a computer program. The computer program is used to cause the processor to execute any one of the above information acquisition methods.

本发明还提供一种计算机程序产品,包括计算机程序,所述计算机程序被处理器执行时实现如上述任一种所述信息获取方法。The present invention also provides a computer program product, which includes a computer program. When the computer program is executed by a processor, it implements any one of the above information acquisition methods.

本发明提供的信息获取方法及系统,通过人脸验证和位置验证的双重身份验证,可以实现用户身份和用户环境的双重安全保障,避免了一些不法用户终端,基于图像处理手段,通过用户身份的验证而造成用户的财产损失。The information acquisition method and system provided by the present invention can achieve dual security guarantees of user identity and user environment through dual identity verification of face verification and location verification, and avoid some unscrupulous user terminals, based on image processing means, through user identity verification. Verification causes property damage to the user.

附图说明Description of the drawings

为了更清楚地说明本发明或现有技术中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作一简单地介绍,显而易见地,下面描述中的附图是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。In order to explain the present invention or the technical solutions in the prior art more clearly, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are of the present invention. For some embodiments of the invention, those of ordinary skill in the art can also obtain other drawings based on these drawings without exerting creative efforts.

图1是本发明提供的信息获取方法的流程示意图;Figure 1 is a schematic flow chart of the information acquisition method provided by the present invention;

图2是本发明提供的卷积神经网络的结构示意图之一;Figure 2 is one of the structural schematic diagrams of the convolutional neural network provided by the present invention;

图3是本发明提供的卷积神经网络的结构示意图之二;Figure 3 is the second structural schematic diagram of the convolutional neural network provided by the present invention;

图4是本发明提供的卷积神经网络中的Inception模块的结构示意图;Figure 4 is a schematic structural diagram of the Inception module in the convolutional neural network provided by the present invention;

图5是本发明提供的信息获取系统的结构示意图;Figure 5 is a schematic structural diagram of the information acquisition system provided by the present invention;

图6是本发明提供的电子设备的实体结构示意图。Figure 6 is a schematic diagram of the physical structure of the electronic device provided by the present invention.

具体实施方式Detailed ways

为使本发明的目的、技术方案和优点更加清楚,下面将结合本发明中的附图,对本发明中的技术方案进行清楚、完整地描述,显然,所描述的实施例是本发明一部分实施例,而不是全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。In order to make the purpose, technical solutions and advantages of the present invention more clear, the technical solutions in the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention. , not all examples. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts fall within the scope of protection of the present invention.

用户在首次进入互联网应用时,需要填写个人信息并绑定手机号码(发送短信验证码)来注册互联网应用账号,在用户正常使用过程中,若涉及到支付款、找回密码等敏感操作,现有互联网应用通常采用校验短信验证码的方式、人脸识别对用户身份进行识别,从而确定是本人操作,这种用户身份验证方式往往无法确保用户支付环境安全,基于此,本发明提供一种信息获取方法,在用户进行身份识别时,加入环境信息、位置信息和人脸表情的多重验证匹配,从而实现身份验证,具体实现如下:When users enter an Internet application for the first time, they need to fill in personal information and bind a mobile phone number (send a SMS verification code) to register an Internet application account. During normal use, if sensitive operations such as payment and password retrieval are involved, now Some Internet applications usually use the method of verifying SMS verification codes and face recognition to identify the user's identity, so as to determine that it is operated by himself. This user identity verification method often cannot ensure the security of the user's payment environment. Based on this, the present invention provides a The information acquisition method adds multiple verification matching of environmental information, location information and facial expressions when the user performs identity recognition to achieve identity verification. The specific implementation is as follows:

图1是本发明提供的信息获取方法的流程示意图,如图1所示,方法包括:Figure 1 is a schematic flow chart of the information acquisition method provided by the present invention. As shown in Figure 1, the method includes:

步骤100、使用第一训练样本对卷积神经网络进行预训练,以获取图像识别模型;Step 100: Use the first training sample to pre-train the convolutional neural network to obtain an image recognition model;

步骤200、使用第二训练样本对图像识别模型进行优化训练,以获取目标图像识别模型;Step 200: Use the second training sample to optimize and train the image recognition model to obtain the target image recognition model;

步骤300、对接收到的用户终端上传的身份验证信息进行人脸验证以及基于目标图像识别模型对身份验证信息进行位置验证;Step 300: Perform face verification on the received identity verification information uploaded by the user terminal and perform location verification on the identity verification information based on the target image recognition model;

步骤400、在人脸验证和位置验证均通过的情况下,将接收到的用户终端发送的信息获取请求对应的数据信息发送给用户终端;Step 400: If both face verification and location verification pass, send the data information corresponding to the received information acquisition request sent by the user terminal to the user terminal;

其中,第一训练样本是根据获取的多数的环境图像中每一环境图像对应的第一位置信息对每一环境图像进行标注后得到的;Wherein, the first training sample is obtained by labeling each environment image according to the first position information corresponding to each environment image in the plurality of acquired environment images;

第二训练样本是根据人脸图像与第一训练样本中的环境图像叠加后的目标图像确定的;The second training sample is determined based on the target image obtained by superimposing the face image and the environment image in the first training sample;

身份验证信息包括用户实时拍摄的包括用户人脸图像的环境图像和用户的位置信息。The identity verification information includes the user's real-time environment image including the user's face image and the user's location information.

需要说明的是,上述方法的执行主体可以是计算机设备。It should be noted that the execution subject of the above method may be a computer device.

可选地,本发明预先将采集的多数的环境图像中每一环境图像对应的位置信息以位置标签的形式对每一环境图像进行标注后得到第一训练样本,并将第一训练样本输入到卷积神经网络中进行训练学习,得到图像识别模型。Optionally, the present invention pre-labels the location information corresponding to each environment image in the plurality of collected environment images in the form of a location label to obtain the first training sample, and inputs the first training sample to Training and learning are performed in the convolutional neural network to obtain the image recognition model.

环境图像是指包括自然场景中的物体的图片,具体可以包括建筑物图片、街景图片、建筑物室内图片等等。在获取环境图像时会同时获取环境图像对应的采集位置信息(即第一位置信息),通过第一位置信息对环境图像进行标注后得到第一训练样本,例如标注内容可以为**城市**街道**路。Environmental images refer to pictures including objects in natural scenes, which can specifically include pictures of buildings, street view pictures, indoor pictures of buildings, etc. When acquiring the environment image, the collection location information corresponding to the environment image (i.e., the first location information) will be obtained at the same time. The first training sample will be obtained after annotating the environment image through the first location information. For example, the annotation content can be **City** Street** Road.

基于图像识别模型,挖掘环境图像与位置信息的关系;然后将人脸图像与第一训练样本中的环境图像进行随机叠加组合后输入到图像识别模型中,得到人脸图像与环境图像对应的位置信息,将位置信息与环境图像对应的位置标签进行比较,优化图像识别模型,以得到目标图像识别模型。Based on the image recognition model, the relationship between the environment image and the location information is mined; then the face image and the environment image in the first training sample are randomly superimposed and combined, and then input into the image recognition model to obtain the corresponding position of the face image and the environment image. Information, compare the location information with the location tag corresponding to the environment image, and optimize the image recognition model to obtain the target image recognition model.

第二训练样本中的人脸图像的环境图像随机叠加是指将人脸图像与环境图像进行自由组合后,将环境图像作为背景图像、人脸图像作为前景图像进行图层叠加。The random superposition of the environment image of the face image in the second training sample means that after freely combining the face image and the environment image, the environment image is used as the background image and the face image is used as the foreground image for layer overlay.

在进行身份验证时,对接收到的用户终端上传的身份验证信息(包括用户实时拍摄的包括用户人脸图像的环境图像和用户的位置信息)进行人脸识别,根据人脸识别结果完成对用户终端上传的身份验证信息的人脸验证,并基于训练得到的目标图像识别模型对身份验证信息进行位置验证。When performing identity verification, face recognition is performed on the received identity verification information uploaded by the user terminal (including the environmental image including the user's face image and the user's location information taken in real time by the user), and the user is completed based on the face recognition results. Face verification of the identity verification information uploaded by the terminal, and location verification of the identity verification information based on the trained target image recognition model.

在进行用户身份验证时,采集包括用户人脸图像的环境图像和用户位置信息的用户的身份验证信息,将包括用户人脸图像的环境图像输入到目标图像识别模型中,得到预测位置信息,然后将预测位置信息与采集到的用户的位置信息进行匹配,完成对用户的位置验证,并结合人脸验证结果,最终实现对用户的位置验证和人脸验证。When performing user identity verification, the user's identity verification information including the environment image of the user's face image and the user's location information is collected, and the environment image including the user's face image is input into the target image recognition model to obtain the predicted location information, and then Match the predicted location information with the collected user location information to complete the user's location verification, and combine it with the face verification results to finally achieve the user's location verification and face verification.

例如,可以采用将包括用户人脸图像的环境图像的人脸特征与预存在特征库中的该用户的人脸特征进行匹配,完成对用户的人脸验证。For example, the face verification of the user can be completed by matching the facial features of the environment image including the user's face image with the user's facial features pre-stored in the feature library.

在人脸验证和位置验证均通过的情况下,将接收到的用户终端发送的信息获取请求对应的数据信息发送给用户终端。When both face verification and location verification pass, the data information corresponding to the received information acquisition request sent by the user terminal is sent to the user terminal.

本发明提供的信息获取方法,通过人脸验证和位置验证的双重身份验证,可以实现用户身份和用户环境的双重安全保障,避免了一些不法用户终端,基于图像处理手段,通过用户身份的验证而造成用户的财产损失。The information acquisition method provided by the present invention can achieve dual security guarantees of user identity and user environment through dual identity verification of face verification and location verification, and avoids the possibility of some unscrupulous user terminals to verify the user identity based on image processing means. Causing property damage to users.

进一步地,在一个实施例中,步骤100,可以具体包括:Further, in one embodiment, step 100 may specifically include:

步骤1001、将第一训练样本输入至卷积神经网络进行预训练,并根据卷积神经网络输出的每一环境图像的第二位置信息与第一位置信息的比较结果,对卷积神经网络的超参数进行调整;Step 1001: Input the first training sample to the convolutional neural network for pre-training, and based on the comparison result of the second position information and the first position information of each environmental image output by the convolutional neural network, perform the convolutional neural network's Hyperparameters are adjusted;

步骤1002、根据调整后的卷积神经网络,确定图像识别模型。Step 1002: Determine the image recognition model based on the adjusted convolutional neural network.

可选地,在步骤1001中,可以通过将得到的第一训练样本输入到如图2所示的卷积神经网络中进行预训练,卷积神经网络具备多层次网络结构,同时也是多层非全连接的神经网络。Optionally, in step 1001, the obtained first training sample can be input into the convolutional neural network as shown in Figure 2 for pre-training. The convolutional neural network has a multi-layer network structure and is also a multi-layer non-linear network. Fully connected neural network.

从功能上来说卷积层和池化层主要是进行图像的特征提取部分,完成图像特征从低维向高维的映射,而全连接层则完成高维特征向图像类别的转变。Functionally speaking, the convolution layer and the pooling layer are mainly responsible for extracting the features of the image and completing the mapping of image features from low dimensions to high dimensions, while the fully connected layer completes the transformation of high-dimensional features into image categories.

可选地,本发明以基于GoogleNet深度卷积网络Recog-Net作为卷积神经网络,其结构如下图3、图4所示:Optionally, the present invention uses Recog-Net, a deep convolutional network based on GoogleNet, as the convolutional neural network. Its structure is shown in Figure 3 and Figure 4 below:

其中,图3为卷积神经网络的整体结构,图4为图3中Inception模块的结构图。在图3中,卷积神经网络的输入为大小为229*229*3的RGB三通道环境图像,而后紧接着三个阶段的卷积操作C1、C2、C3,接着伴随的p1池化降维以及C4、C5的卷积以提取更为抽象的高层特征。Among them, Figure 3 shows the overall structure of the convolutional neural network, and Figure 4 shows the structure diagram of the Inception module in Figure 3. In Figure 3, the input of the convolutional neural network is an RGB three-channel environment image with a size of 229*229*3, followed by three stages of convolution operations C1, C2, and C3, followed by the accompanying p1 pooling dimensionality reduction. and convolution of C4 and C5 to extract more abstract high-level features.

如图4所示,卷积神经网络的核心为Inception模块,整个inception模块是由1*1,3*3和5*5三个不同尺度的卷积核组成。同时配合池化操作(3*3)堆叠在一起(卷积、池化后的尺寸相同,将通道相加),一方面增加了卷积神经网络的宽度,另一方面也增加了卷积神经网络对尺度的契合性。在每一个卷积层后都要做一个ReLU操作,ReLU函数作为一个非线性激活函数可以有效地拟合卷积神经网络的训练状态。As shown in Figure 4, the core of the convolutional neural network is the Inception module. The entire inception module is composed of three convolution kernels of different scales: 1*1, 3*3 and 5*5. At the same time, it is stacked together with the pooling operation (3*3) (the sizes after convolution and pooling are the same, and the channels are added). On the one hand, it increases the width of the convolutional neural network, and on the other hand, it also increases the number of convolutional neural networks. Network fit to scale. A ReLU operation is performed after each convolutional layer. As a nonlinear activation function, the ReLU function can effectively fit the training state of the convolutional neural network.

在预训练阶段,将卷积神经网络的预测结果与对应第一训练样本标注的第一位置信息进行比较,然后根据比较结果对卷积神经网络进行优化,具体地,可以通过对卷积神经网络的超参数进行优化调整,从而确定调整后的超参数,并根据调整后的卷积神经网络,得到图像识别模型。In the pre-training stage, the prediction results of the convolutional neural network are compared with the first position information marked corresponding to the first training sample, and then the convolutional neural network is optimized based on the comparison results. Specifically, the convolutional neural network can be optimized by Optimize and adjust the hyperparameters to determine the adjusted hyperparameters, and obtain the image recognition model based on the adjusted convolutional neural network.

本发明提供的信息获取方法,使用第一训练样本训练卷积神经网络,以获取图像识别模型,为后续基于图像识别网络获取目标识别模型,并最终实现对用户位置验证奠定了基础。The information acquisition method provided by the present invention uses the first training sample to train the convolutional neural network to obtain the image recognition model, which lays the foundation for subsequent acquisition of the target recognition model based on the image recognition network and ultimately the verification of the user's location.

进一步地,在一个实施例中,步骤200可以具体包括:Further, in one embodiment, step 200 may specifically include:

步骤2001、将第二训练样本输入至图像识别模型进行优化训练,并根据图像识别模型输出的目标图像的第三位置信息与第一位置信息的比较结果,采用反向传播算法和随机梯度算法对图像识别模型进行优化,以获取目标图像识别模型。Step 2001: Input the second training sample to the image recognition model for optimization training, and use the back propagation algorithm and the stochastic gradient algorithm to compare the third position information and the first position information of the target image output by the image recognition model. The image recognition model is optimized to obtain the target image recognition model.

可选地,将第二训练样本输入到步骤100预训练得到的图像识别模型中,将图像识别模型预测得到的位置数据(即第三位置信息)与第二训练样本中标注的环境图像的位置信息(与第一位置信息一致)进行比较,并根据比较结果采用连续的反向传播算法结合动态改变学习率的随机梯度下降算法来优化训练图像识别模型,得到优化后的图像识别模型,并将优化后的图像识别模型作为目标图像识别模型。Optionally, the second training sample is input into the image recognition model pre-trained in step 100, and the position data predicted by the image recognition model (ie, the third position information) is combined with the position of the environment image marked in the second training sample. information (consistent with the first position information), and based on the comparison results, a continuous backpropagation algorithm combined with a stochastic gradient descent algorithm that dynamically changes the learning rate is used to optimize the training of the image recognition model, and the optimized image recognition model is obtained, and The optimized image recognition model is used as the target image recognition model.

例如,本发明中的环境图像的位置识别可以是针对建筑物图像的位置识别,为了让使训练得到的目标图像识别模型拥有更加准确的判别能力,本发明通过原始的建筑物图像作为第一训练样本进行卷积神经网络预训练,使得训练得到的图像识别模型能够精准的判断建筑物的位置;然后,将人脸图像与建筑物图像进行随机叠加处理,将得到的随机叠加后的图像作为第二训练样本对预训练后得到的图像识别模型进行二次优化训练,可以使得经过两次训练后得到的目标图像识别模型能够基于预训练的图像识别模型记忆实现对包括人脸图像的建筑物图像进行位置的准确预测。For example, the location recognition of the environment image in the present invention can be the location recognition of the building image. In order to make the trained target image recognition model have more accurate discrimination ability, the present invention uses the original building image as the first training The sample is pre-trained with a convolutional neural network so that the trained image recognition model can accurately determine the location of the building; then, the face image and the building image are randomly superimposed, and the resulting randomly superimposed image is used as the third Second training samples are used to perform secondary optimization training on the image recognition model obtained after pre-training, so that the target image recognition model obtained after two trainings can realize building images including face images based on the memory of the pre-trained image recognition model. Make accurate predictions of location.

本发明提供的信息获取方法,使用第二训练样本训练训练图像识别模型,为后续基于训练好的图像识别网络确定目标识别模型,并最终实现对用户位置验证奠定了基础。The information acquisition method provided by the present invention uses the second training sample to train the image recognition model, which lays the foundation for subsequent determination of the target recognition model based on the trained image recognition network, and ultimately the verification of the user's location.

进一步地,在一个实施例中,步骤300中的用户终端上传的身份验证信息进行位置是通过如下方式确定的:Further, in one embodiment, the location of the identity verification information uploaded by the user terminal in step 300 is determined in the following manner:

在接收到用户终端发送的信息获取请求后,向用户终端发送身份验证信息采集信息,以使用户终端根据所述身份验证信息采集信息,采集身份验证信息。After receiving the information acquisition request sent by the user terminal, identity verification information collection information is sent to the user terminal, so that the user terminal collects information according to the identity verification information and collects identity verification information.

可选地,在众多的互联网应用中,如微信、支付宝、美团、滴滴等,尤其像移动的5G消息功能,实现5G消息相关应用功能,均会涉及到用户身份验证。目前的身份验证主要是通过人脸识别、指纹识别等实现。Optionally, in many Internet applications, such as WeChat, Alipay, Meituan, Didi, etc., especially mobile 5G messaging functions, implementing 5G messaging related application functions will involve user identity verification. Current identity verification is mainly achieved through face recognition, fingerprint recognition, etc.

本发明则提出了一种新的身份验证方式,结合环境图像的位置身份验证。用户在使用互联网应用时会通过以发送信息获取请求的方式从服务端获取目标页面信息,如个人中心页面、密码支付页面等。在信息获取请求中,则会包括目标请求信息的地址数据、用户的身份信息(如用户终端标识、用户账号或号码等)。The present invention proposes a new identity verification method that combines location identity verification with environmental images. When users use Internet applications, they will obtain target page information from the server by sending information acquisition requests, such as personal center pages, password payment pages, etc. In the information acquisition request, the address data of the target request information and the user's identity information (such as user terminal identification, user account or number, etc.) will be included.

服务端基于信息获取请求中包括的用户身份信息,向用户终端发送身份验证信息采集信息。The server sends identity verification information to the user terminal to collect information based on the user identity information included in the information acquisition request.

用户终端在接收到身份验证信息采集信息后,通过采集装置拍摄包括用户人脸图像的环境图像和用户的位置信息,作为身份验证信息上传到服务端,服务端接收身份验证信息,获取环境图像;同时身份验证信息中包括用户身份信息,如用户手机号。After receiving the identity verification information collection information, the user terminal captures the environment image including the user's face image and the user's location information through the collection device, and uploads it to the server as identity verification information. The server receives the identity verification information and obtains the environment image; At the same time, the identity verification information includes user identity information, such as user mobile phone number.

该采集装置可以具体为内置的摄像头采集包括用户人脸图像的环境图像和用户的位置信息。也可以具体为通过外置的、且与用户终端关联的摄像头采集包括用户人脸图像的环境图像和用户的位置信息。比如,用户终端可通过连接线或网络与图像采集设备连接,图像采集设备通过摄像头采集包括用户人脸图像的环境图像和用户的位置信息,并将采集包括用户人脸图像的环境图像和用户的位置信息传输至用户终端。摄像头可以是单目摄像头、双目摄像头、深度摄像头、3D(3 Dimensions,三维)摄像头等。用户终端可采集现实场景中活体的图像,也可对现实场景中包含人脸的已有图像进行采集,比如身份证件扫描件等。The collection device may specifically be a built-in camera that collects environmental images including user face images and user location information. It can also specifically collect the environment image including the user's face image and the user's location information through an external camera associated with the user terminal. For example, the user terminal can be connected to the image acquisition device through a connecting line or network. The image acquisition device collects the environment image including the user's face image and the user's location information through the camera, and will collect the environment image including the user's face image and the user's location information. The location information is transmitted to the user terminal. The camera can be a monocular camera, a binocular camera, a depth camera, a 3D (3 Dimensions, three-dimensional) camera, etc. The user terminal can collect images of living objects in real scenes, and can also collect existing images containing faces in real scenes, such as scans of ID documents, etc.

其中,用户终端可以但不限于是各种智能手机、平板电脑、笔记本电脑、台式计算机、便携式可穿戴设备、智能音箱等。服务端可以具体为服务器,服务器可以是独立的物理服务器,或者是多个物理服务器构成的服务器集群或者分布式系统,或者提供云服务、云数据库、云计算、云函数、存储、网络服务、云通信、中间件服务、域名服务、安全服务、CDN、以及大数据和人工智能平台等基础云计算服务的云服务器。Among them, the user terminal can be, but is not limited to, various smart phones, tablets, laptops, desktop computers, portable wearable devices, smart speakers, etc. The server can be specifically a server. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or it can provide cloud services, cloud databases, cloud computing, cloud functions, storage, network services, and cloud services. Cloud servers for basic cloud computing services such as communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms.

本发明提供的信息获取方法,通过对用户终端上传的身份验证信息中包括的用户实时拍摄的包括用户人脸图像的环境图像和用户的位置信息,实现对用户的人脸验证和位置验证的双重身份验证,提高了验证的安全性。The information acquisition method provided by the present invention realizes the dual face verification and location verification of the user by collecting the user's real-time environment image including the user's face image and the user's location information included in the identity verification information uploaded by the user terminal. Identity verification improves the security of verification.

进一步地,在一个实施例中,步骤400可以具体包括:Further, in one embodiment, step 400 may specifically include:

步骤4001、提取包括用户人脸图像的环境图像的第一人脸特征;Step 4001: Extract the first facial feature of the environment image including the user's facial image;

步骤4002、将第一人脸特征与身份验证信息中用户注册时的第二人脸特征进行特征向量计算,以获取第一人脸特征与第二人脸特征之间的距离;Step 4002: Calculate the feature vector between the first facial feature and the second facial feature in the identity verification information when the user is registered to obtain the distance between the first facial feature and the second facial feature;

步骤4003、根据距离,完成对身份验证信息的人脸验证;Step 4003: Complete the face verification of the identity verification information based on the distance;

步骤4004、在人脸验证通过的情况下,基于目标图像识别模型对所述身份验证信息进行位置验证。Step 4004: If the face verification passes, perform location verification on the identity verification information based on the target image recognition model.

可选地,根据用户终端上传的身份验证信息中的包括用户人脸图像的环境图像,由服务端通过图像处理模块,从该环境图像中提取人脸特征,即第一人脸特征。Optionally, according to the environment image including the user's face image in the identity verification information uploaded by the user terminal, the server extracts the facial features, that is, the first facial features, from the environment image through the image processing module.

通过身份验证信息中的用户身份信息,从人脸特征库中获取用户注册时得到的人脸特征,即第二人脸特征。Through the user identity information in the identity verification information, the facial features obtained when the user registered, that is, the second facial features, are obtained from the facial feature database.

人脸特征是人脸所固有的生理特征,比如,虹膜形态、面部器官(眼睛、鼻子、嘴、耳朵等)之间的位置关系、面部器官的结构(形状、大小等)、皮肤纹理等等。Facial features are the inherent physiological characteristics of the human face, such as iris shape, positional relationship between facial organs (eyes, nose, mouth, ears, etc.), structure of facial organs (shape, size, etc.), skin texture, etc. .

将从用户终端上传的身份验证信息中的包括用户人脸图像的环境图像中提取的第一人脸特征与人脸特征库得到的第二人脸特征进行特征向量计算,例如可以通过马氏距离得到第一人脸特征与第二人脸特征之间的距离,完成对身份验证信息中包括用户人脸图像的环境图像的人脸验证,若计算得到的该距离为零,则人脸验证通过。Perform feature vector calculation on the first facial feature extracted from the environment image including the user's face image in the identity verification information uploaded by the user terminal and the second facial feature obtained from the face feature library. For example, the Mahalanobis distance can be used Obtain the distance between the first facial feature and the second facial feature, and complete the face verification of the environment image including the user's face image in the identity verification information. If the calculated distance is zero, the face verification is passed. .

在人脸验证通过的情况下,基于目标图像识别模型对身份验证信息中用户的位置信息进行位置验证。When the face verification is passed, the location verification of the user's location information in the identity verification information is performed based on the target image recognition model.

本发明提供的信息获取方法,通过预测环境图像所对应的位置信息,将预测得到的位置信息与用户采集的位置信息进行比较,实现基于位置信息的环境图像验证,提高了验证的安全性。The information acquisition method provided by the present invention realizes environmental image verification based on location information by predicting the location information corresponding to the environment image and comparing the predicted location information with the location information collected by the user, thereby improving the security of the verification.

进一步地,在一个实施例中,步骤4004可以具体包括:Further, in one embodiment, step 4004 may specifically include:

步骤40041、将包括用户人脸图像的环境图像输入至目标图像识别模型,并根据目标图像识别模型输出的第四位置信息与用户的位置信息的比较结果,完成对身份验证信息的位置验证。Step 40041: Input the environment image including the user's face image to the target image recognition model, and complete the location verification of the identity verification information based on the comparison result between the fourth location information output by the target image recognition model and the user's location information.

可选地,将包括用户人脸图像的环境图像输入至目标图像识别模型,并将目标图像识别模型输出的包括人脸特征的环境图像的位置信息(即第四位置信息)进行比较,以完成对用户的位置信息的位置验证。Optionally, input the environment image including the user's face image to the target image recognition model, and compare the position information (ie, the fourth position information) of the environment image including the face features output by the target image recognition model to complete Location verification of the user's location information.

例如,位置比较可以具体为,将采集的用户的位置信息Gj与预测得到的第四位置信息Gi进行比较,若采集的用户的位置信息Gj位于预测的第四位置信息Gi的地理区间规划范围内,则可以通过位置验证。在进行位置验证的过程中,除了简单的进行地理区间范围规划外,在采集用户的位置信息的过程中,利用其他技术保证采集到的用户的位置信息的真实性。For example, the location comparison may specifically include comparing the collected user's location information G j with the predicted fourth location information G i . If the collected user's location information G j is located in the geographical location of the predicted fourth location information G i Within the interval planning range, the location verification can be passed. In the process of location verification, in addition to simple geographical interval planning, other technologies are used to ensure the authenticity of the collected user location information during the process of collecting the user's location information.

例如,通过查看地图软件、微信等用户的实时位置与用户终端上传的身份验证信息中的用户的位置信息区别,以保证采集到的用户的位置信息的真实性。For example, by checking the difference between the user's real-time location on map software, WeChat, etc. and the user's location information in the identity verification information uploaded by the user terminal, the authenticity of the collected user's location information can be ensured.

又比如,可以引入区块链技术,将服务端作为区块链网络中的区块链节点,通过将包括用户实时拍摄的包括用户人脸图像的环境图像和用户的位置信息存储在区块链上,并从区块链的数据区块中获取用户的位置信息,以确保用户终端上传的用户的位置信息与区块链中存储的用户的位置信息的一致。区块链是分布式数据存储、点对点传输、共识机制、加密算法等计算机技术的新型应用模式。区块链(Blockchain),本质上是一个去中心化的数据库,是一串使用密码学方法相关联产生的数据块,每一个数据块中包含了一批次网络交易的信息,用于验证其信息的有效性(防伪)和生成下一个区块。区块链可以包括区块链底层平台、平台产品服务层以及应用服务层。For another example, blockchain technology can be introduced, and the server can be used as a blockchain node in the blockchain network, and the environmental images including the user's face image and the user's location information captured in real time by the user can be stored in the blockchain. and obtain the user's location information from the data block of the blockchain to ensure that the user's location information uploaded by the user terminal is consistent with the user's location information stored in the blockchain. Blockchain is a new application model of computer technology such as distributed data storage, point-to-point transmission, consensus mechanism, and encryption algorithm. Blockchain is essentially a decentralized database. It is a series of data blocks generated using cryptographic methods. Each data block contains a batch of network transaction information and is used to verify its Validity of information (anti-counterfeiting) and generation of the next block. Blockchain can include the underlying platform of the blockchain, the platform product service layer and the application service layer.

其中,Q=1表示通过位置验证,Q=-1表示不通过位置验证。Among them, Q=1 indicates that the location verification is passed, and Q=-1 indicates that the location verification is not passed.

在人脸验证通过和位置验证通过时,发送信息获取请求对应的数据信息,具体地:When the face verification and location verification pass, a message is sent to obtain the data information corresponding to the request, specifically:

在人脸验证通过、位置验证未通过时,则不通过身份验证;If the face verification is passed but the location verification is not passed, the identity verification will not be passed;

在人脸验证未通过时,则不进行位置验证,可以节省验证的处理时间;When face verification fails, location verification is not performed, which can save verification processing time;

在人脸验证通过、位置验证通过时,则通过身份验证,发送接收到的用户终端发送的信息获取请求对应的数据信息。When the face verification is passed and the location verification is passed, the data information corresponding to the received information acquisition request sent by the user terminal is sent through identity verification.

在实际进行用户身份验证时,还可以先对用户终端上传的身份验证信息中用户的位置信息进行位置验证,并在位置验证未通过时,不进行人脸验证,以节省验证的处理时间。When actually performing user identity verification, you can also perform location verification on the user's location information in the identity verification information uploaded by the user terminal, and when the location verification fails, face verification is not performed to save verification processing time.

本发明提供的信息获取方法,通过人脸验证和位置验证的双重身份验证,可以实现用户身份和用户环境的双重安全保障,可以保证用户在安全的环境下进行支付等操作,避免了一些不法用户终端,基于图像处理手段,通过用户身份的验证而造成用户的财产损失。The information acquisition method provided by the present invention can achieve dual security guarantees of user identity and user environment through dual identity verification of face verification and location verification, and can ensure that users perform operations such as payment in a safe environment, avoiding some unscrupulous users. The terminal, based on image processing means, causes the user's property loss through user identity verification.

下面对本发明提供的信息获取系统进行描述,下文描述的信息获取系统与上文描述的信息获取方法可相互对应参照。The information acquisition system provided by the present invention is described below. The information acquisition system described below and the information acquisition method described above can be referenced correspondingly.

图5是本发明提供的信息获取系统的结构示意图,如图5所示,包括:Figure 5 is a schematic structural diagram of the information acquisition system provided by the present invention. As shown in Figure 5, it includes:

第一训练模块510、第二训练模块511、身份验证模块512和信息获取模块513;The first training module 510, the second training module 511, the identity verification module 512 and the information acquisition module 513;

第一训练模块510,用于使用第一训练样本对卷积神经网络进行预训练,以获取图像识别模型;The first training module 510 is used to pre-train the convolutional neural network using the first training sample to obtain the image recognition model;

第二训练模块511,用于使用第二训练样本对图像识别模型进行优化训练,以获取目标图像识别模型;The second training module 511 is used to optimize and train the image recognition model using the second training sample to obtain the target image recognition model;

身份验证模块512,用于对接收到的用户终端上传的身份验证信息进行人脸验证以及基于目标图像识别模型对身份验证信息进行位置验证;The identity verification module 512 is used to perform face verification on the received identity verification information uploaded by the user terminal and perform location verification on the identity verification information based on the target image recognition model;

信息获取模块513,用于在人脸验证和位置验证均通过的情况下,将接收到的用户终端发送的信息获取请求对应的数据信息发送给用户终端;The information acquisition module 513 is used to send the data information corresponding to the received information acquisition request sent by the user terminal to the user terminal when both the face verification and the location verification are passed;

其中,第一训练样本是根据获取的多数的环境图像中每一环境图像对应的第一位置信息对每一环境图像进行标注后得到的;Wherein, the first training sample is obtained by labeling each environment image according to the first position information corresponding to each environment image in the plurality of acquired environment images;

第二训练样本是根据人脸图像与第一训练样本中的环境图像叠加后的目标图像确定的;The second training sample is determined based on the target image obtained by superimposing the face image and the environment image in the first training sample;

身份验证信息包括用户实时拍摄的包括用户人脸图像的环境图像和用户的位置信息。The identity verification information includes the user's real-time environment image including the user's face image and the user's location information.

本发明提供的信息获取系统,通过人脸验证和位置验证的双重身份验证,可以实现用户身份和用户环境的双重安全保障,避免了一些不法用户终端,基于图像处理手段,通过用户身份的验证而造成用户的财产损失。The information acquisition system provided by the present invention can achieve dual security guarantees of user identity and user environment through dual identity verification of face verification and location verification, and avoids the possibility of some unscrupulous user terminals to verify the user identity based on image processing means. Causing property damage to users.

进一步地,在一个实施例中,第一训练模块510,还可以具体用于:Further, in one embodiment, the first training module 510 can also be specifically used for:

将第一训练样本输入至卷积神经网络进行预训练,并根据卷积神经网络输出的每一环境图像的第二位置信息与第一位置信息的比较结果,对卷积神经网络的超参数进行调整;Input the first training sample to the convolutional neural network for pre-training, and perform the hyperparameters of the convolutional neural network based on the comparison results between the second position information and the first position information of each environmental image output by the convolutional neural network. Adjustment;

根据调整后的卷积神经网络,确定图像识别模型。Based on the adjusted convolutional neural network, the image recognition model is determined.

本发明提供的信息获取系统,使用第一训练样本训练卷积神经网络,以获取图像识别模型,为后续基于图像识别网络获取目标识别模型,并最终实现对用户位置验证奠定了基础。The information acquisition system provided by the present invention uses the first training sample to train the convolutional neural network to obtain the image recognition model, which lays the foundation for subsequent acquisition of the target recognition model based on the image recognition network and ultimately the verification of the user's location.

进一步地,在一个实施例中,第二训练模块511,还可以具体用于:Further, in one embodiment, the second training module 511 can also be specifically used for:

将第二训练样本输入至图像识别模型进行优化训练,并根据图像识别模型输出的目标图像的第三位置信息与第一位置信息的比较结果,采用反向传播算法和随机梯度算法对图像识别模型进行优化,以获取目标图像识别模型。The second training sample is input to the image recognition model for optimization training, and based on the comparison result between the third position information and the first position information of the target image output by the image recognition model, the back propagation algorithm and the stochastic gradient algorithm are used to optimize the image recognition model Optimize to obtain the target image recognition model.

本发明提供的信息获取系统,使用第二训练样本训练训练图像识别模型,为后续基于训练好的图像识别网络确定目标识别模型,并最终实现对用户位置验证奠定了基础。The information acquisition system provided by the present invention uses the second training sample to train the image recognition model, which lays the foundation for subsequent determination of the target recognition model based on the trained image recognition network, and ultimately the verification of the user's location.

进一步地,在一个实施例中,身份验证模块512,还可以具体用于:Further, in one embodiment, the identity verification module 512 can also be specifically used to:

在接收到用户终端发送的信息获取请求后,向用户终端发送身份验证信息采集信息,以使用户终端根据所述身份验证信息采集信息对身份验证信息进行采集。After receiving the information acquisition request sent by the user terminal, identity verification information collection information is sent to the user terminal, so that the user terminal collects identity verification information based on the identity verification information collection information.

本发明提供的信息获取系统,通过对用户终端上传的身份验证信息中包括的用户实时拍摄的包括用户人脸图像的环境图像和用户的位置信息,实现对用户的人脸验证和位置验证的双重身份验证,提高了验证的安全性。The information acquisition system provided by the present invention realizes the dual face verification and location verification of the user by collecting the user's real-time environment image including the user's face image and the user's location information included in the identity verification information uploaded by the user terminal. Identity verification improves the security of verification.

进一步地,在一个实施例中,身份验证模块512,还可以具体用于:Further, in one embodiment, the identity verification module 512 can also be specifically used to:

提取包括用户人脸图像的环境图像的第一人脸特征;extracting a first facial feature of an environment image including a user's facial image;

将第一人脸特征与身份验证信息中用户注册时的第二人脸特征进行特征向量计算,以获取第一人脸特征与第二人脸特征之间的距离;Calculate feature vectors between the first facial feature and the second facial feature when the user is registered in the identity verification information to obtain the distance between the first facial feature and the second facial feature;

根据距离,完成对身份验证信息的人脸验证;Complete face verification of identity verification information based on distance;

在人脸验证通过的情况下,基于目标图像识别模型对身份验证信息进行位置验证。When the face verification is passed, the identity verification information is location verified based on the target image recognition model.

本发明提供的信息获取系统,通过预测环境图像所对应的位置信息,将预测得到的位置信息与用户采集的位置信息进行比较,实现基于位置信息的环境图像验证,提高了验证的安全性。The information acquisition system provided by the present invention realizes environmental image verification based on location information by predicting the location information corresponding to the environment image and comparing the predicted location information with the location information collected by the user, thereby improving the security of the verification.

进一步地,在一个实施例中,身份验证模块512,还可以具体用于:Further, in one embodiment, the identity verification module 512 can also be specifically used to:

将包括用户人脸图像的环境图像输入至目标图像识别模型,并根据目标图像识别模型输出的第四位置信息与用户的位置信息的比较结果,完成对身份验证信息的位置验证。The environment image including the user's face image is input to the target image recognition model, and the location verification of the identity verification information is completed based on the comparison result between the fourth location information output by the target image recognition model and the user's location information.

本发明提供的信息获取系统,通过人脸验证和位置验证的双重身份验证,可以实现用户身份和用户环境的双重安全保障,可以保证用户在安全的环境下进行支付等操作,避免了一些不法用户终端,基于图像处理手段,通过用户身份的验证而造成用户的财产损失。The information acquisition system provided by the present invention can achieve dual security guarantees of user identity and user environment through dual identity verification of face verification and location verification, and can ensure that users perform operations such as payment in a safe environment, avoiding some unscrupulous users. The terminal, based on image processing means, causes the user's property loss through user identity verification.

图6是本发明提供的一种电子设备的实体结构示意图,如图6所示,该电子设备可以包括:处理器(processor)610、通信接口(communication interface)611、存储器(memory)612和总线(bus)613,其中,处理器610,通信接口611,存储器612通过总线613完成相互间的通信。处理器610可以调用存储器612中的逻辑指令,以执行如下方法:Figure 6 is a schematic diagram of the physical structure of an electronic device provided by the present invention. As shown in Figure 6, the electronic device may include: a processor (processor) 610, a communication interface (communication interface) 611, a memory (memory) 612 and a bus. (bus) 613, in which the processor 610, the communication interface 611, and the memory 612 complete communication with each other through the bus 613. Processor 610 can call logic instructions in memory 612 to perform the following methods:

使用第一训练样本对卷积神经网络进行预训练,以获取图像识别模型;Pre-train the convolutional neural network using the first training sample to obtain the image recognition model;

使用第二训练样本对图像识别模型进行优化训练,以获取目标图像识别模型;Use the second training sample to optimize and train the image recognition model to obtain the target image recognition model;

对接收到的用户终端上传的身份验证信息进行人脸验证以及基于目标图像识别模型对身份验证信息进行位置验证;Perform face verification on the received identity verification information uploaded by the user terminal and perform location verification on the identity verification information based on the target image recognition model;

在人脸验证和位置验证均通过的情况下,将接收到的用户终端发送的信息获取请求对应的数据信息发送给用户终端;When both face verification and location verification pass, the data information corresponding to the received information acquisition request sent by the user terminal is sent to the user terminal;

其中,第一训练样本是根据获取的多数的环境图像中每一环境图像对应的第一位置信息对每一环境图像进行标注后得到的;Wherein, the first training sample is obtained by labeling each environment image according to the first position information corresponding to each environment image in the plurality of acquired environment images;

第二训练样本是根据人脸图像与第一训练样本中的环境图像叠加后的目标图像确定的;The second training sample is determined based on the target image obtained by superimposing the face image and the environment image in the first training sample;

身份验证信息包括用户实时拍摄的包括用户人脸图像的环境图像和用户的位置信息。The identity verification information includes the user's real-time environment image including the user's face image and the user's location information.

此外,上述的存储器中的逻辑指令可以通过软件功能单元的形式实现并作为独立的产品销售或使用时,可以存储在一个计算机可读取存储介质中。基于这样的理解,本发明的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的部分可以以软件产品的形式体现出来,该计算机软件产品存储在一个存储介质中,包括若干指令用以使得一台计算机电源屏(可以是个人计算机,服务器,或者网络电源屏等)执行本发明各个实施例所述方法的全部或部分步骤。而前述的存储介质包括:U盘、移动硬盘、只读存储器(ROM,Read-only Memory)、随机存取存储器(RAM,Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。In addition, the above-mentioned logical instructions in the memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention essentially or the part that contributes to the existing technology or the part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer power panel (which can be a personal computer, a server, or a network power panel, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code. .

进一步地,本发明公开一种计算机程序产品,所述计算机程序产品包括存储在非暂态计算机可读存储介质上的计算机程序,所述计算机程序包括程序指令,当所述程序指令被计算机执行时,计算机能够执行上述各方法实施例所提供的信息获取方法,例如包括:Further, the present invention discloses a computer program product. The computer program product includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer , the computer can execute the information acquisition method provided by each of the above method embodiments, including, for example:

使用第一训练样本对卷积神经网络进行预训练,以获取图像识别模型;Pre-train the convolutional neural network using the first training sample to obtain the image recognition model;

使用第二训练样本对图像识别模型进行优化训练,以获取目标图像识别模型;Use the second training sample to optimize and train the image recognition model to obtain the target image recognition model;

对接收到的用户终端上传的身份验证信息进行人脸验证以及基于目标图像识别模型对身份验证信息进行位置验证;Perform face verification on the received identity verification information uploaded by the user terminal and perform location verification on the identity verification information based on the target image recognition model;

在人脸验证和位置验证均通过的情况下,将接收到的用户终端发送的信息获取请求对应的数据信息发送给用户终端;When both face verification and location verification pass, the data information corresponding to the received information acquisition request sent by the user terminal is sent to the user terminal;

其中,第一训练样本是根据获取的多数的环境图像中每一环境图像对应的第一位置信息对每一环境图像进行标注后得到的;Wherein, the first training sample is obtained by labeling each environment image according to the first position information corresponding to each environment image in the plurality of acquired environment images;

第二训练样本是根据人脸图像与第一训练样本中的环境图像叠加后的目标图像确定的;The second training sample is determined based on the target image obtained by superimposing the face image and the environment image in the first training sample;

身份验证信息包括用户实时拍摄的包括用户人脸图像的环境图像和用户的位置信息。The identity verification information includes the user's real-time environment image including the user's face image and the user's location information.

另一方面,本发明还提供一种处理器可读存储介质,所述处理器可读存储介质存储有计算机程序,所述计算机程序用于使所述处理器执行上述各实施例提供的方法,例如包括On the other hand, the present invention also provides a processor-readable storage medium. The processor-readable storage medium stores a computer program. The computer program is used to cause the processor to execute the methods provided in the above embodiments. Examples include

使用第一训练样本对卷积神经网络进行预训练,以获取图像识别模型;Pre-train the convolutional neural network using the first training sample to obtain the image recognition model;

使用第二训练样本对图像识别模型进行优化训练,以获取目标图像识别模型;Use the second training sample to optimize and train the image recognition model to obtain the target image recognition model;

对接收到的用户终端上传的身份验证信息进行人脸验证以及基于目标图像识别模型对身份验证信息进行位置验证;Perform face verification on the received identity verification information uploaded by the user terminal and perform location verification on the identity verification information based on the target image recognition model;

在人脸验证和位置验证均通过的情况下,将接收到的用户终端发送的信息获取请求对应的数据信息发送给用户终端;When both face verification and location verification pass, the data information corresponding to the received information acquisition request sent by the user terminal is sent to the user terminal;

其中,第一训练样本是根据获取的多数的环境图像中每一环境图像对应的第一位置信息对每一环境图像进行标注后得到的;Wherein, the first training sample is obtained by labeling each environment image according to the first position information corresponding to each environment image in the plurality of acquired environment images;

第二训练样本是根据人脸图像与第一训练样本中的环境图像叠加后的目标图像确定的;The second training sample is determined based on the target image obtained by superimposing the face image and the environment image in the first training sample;

身份验证信息包括用户实时拍摄的包括用户人脸图像的环境图像和用户的位置信息。The identity verification information includes the user's real-time environment image including the user's face image and the user's location information.

以上所描述的系统实施例仅仅是示意性的,其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的,作为单元显示的部件可以是或者也可以不是物理单元,即可以位于一个地方,或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。本领域普通技术人员在不付出创造性的劳动的情况下,即可以理解并实施。The system embodiments described above are only illustrative. The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in One location, or it can be distributed across multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Persons of ordinary skill in the art can understand and implement the method without any creative effort.

通过以上的实施方式的描述,本领域的技术人员可以清楚地了解到各实施方式可借助软件加必需的通用硬件平台的方式来实现,当然也可以通过硬件。基于这样的理解,上述技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来,该计算机软件产品可以存储在计算机可读存储介质中,如ROM/RAM、磁碟、光盘等,包括若干指令用以使得一台计算机电源屏(可以是个人计算机,服务器,或者网络电源屏等)执行各个实施例或者实施例的某些部分所述的方法。Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the part of the above technical solution that essentially contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., including a number of instructions to cause a computer power panel (which can be a personal computer, a server, or a network power panel, etc.) to execute the methods described in various embodiments or certain parts of the embodiments.

最后应说明的是:以上实施例仅用以说明本发明的技术方案,而非对其限制;尽管参照前述实施例对本发明进行了详细的说明,本领域的普通技术人员应当理解:其依然可以对前述各实施例所记载的技术方案进行修改,或者对其中部分技术特征进行等同替换;而这些修改或者替换,并不使相应技术方案的本质脱离本发明各实施例技术方案的精神和范围。Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that it can still be used Modifications are made to the technical solutions described in the foregoing embodiments, or equivalent substitutions are made to some of the technical features; however, these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (10)

1. An information acquisition method, characterized by comprising:
pre-training the convolutional neural network by using a first training sample to obtain an image recognition model;
performing optimization training on the image recognition model by using a second training sample to obtain a target image recognition model;
carrying out face verification on the received identity verification information uploaded by the user terminal, and carrying out position verification on the identity verification information based on the target image recognition model;
under the condition that the face verification and the position verification are both passed, the received data information corresponding to the information acquisition request sent by the user terminal is sent to the user terminal;
The first training sample is obtained by marking each environment image according to first position information corresponding to each environment image in a plurality of acquired environment images;
the second training sample is determined according to a target image obtained by overlapping the face image and the environment image in the first training sample;
the identity verification information comprises an environment image including a face image of the user and position information of the user, wherein the environment image is shot by the user in real time.
2. The method of claim 1, wherein the pre-training the convolutional neural network using the first training sample to obtain the image recognition model comprises:
inputting the first training sample into the convolutional neural network for pre-training, and adjusting the super-parameters of the convolutional neural network according to the comparison result of the second position information and the first position information of each environment image output by the convolutional neural network;
and determining the image recognition model according to the adjusted convolutional neural network.
3. The information acquisition method according to claim 1, wherein the optimally training the image recognition model using the second training sample to acquire a target image recognition model includes:
And inputting the second training sample into the image recognition model for optimization training, and optimizing the image recognition model by adopting a back propagation algorithm and a random gradient algorithm according to a comparison result of the third position information of the target image and the first position information output by the image recognition model so as to acquire the target image recognition model.
4. The information acquisition method according to claim 1, wherein the authentication information processing position uploaded by the user terminal is determined by:
after receiving the information acquisition request sent by the user terminal, sending authentication information acquisition information to the user terminal, so that the user terminal acquires the authentication information according to the received authentication information acquisition information.
5. The method of claim 1, wherein the performing face verification on the received authentication information uploaded by the user terminal and performing location verification on the authentication information based on the target image recognition model includes:
extracting first face features of the environment image comprising the face image of the user;
Performing feature vector calculation on the first face feature and a second face feature when a user registers in the identity verification information to obtain a distance between the first face feature and the second face feature;
according to the distance, finishing face verification of the identity verification information;
and under the condition that the face verification is passed, carrying out position verification on the identity verification information based on the target image recognition model.
6. The information acquisition method according to claim 5, wherein the performing location verification of the authentication information based on the target image recognition model includes:
and inputting the environment image comprising the face image of the user into the target image recognition model, and completing the position verification of the identity verification information according to the comparison result of the fourth position information output by the target image recognition model and the position information of the user.
7. An information acquisition system, comprising: the system comprises a first training module, a second training module, an identity verification module and an information acquisition module;
the first training module is used for pre-training the convolutional neural network by using a first training sample so as to obtain an image recognition model;
The second training module is used for carrying out optimization training on the image recognition model by using a second training sample so as to obtain a target image recognition model;
the identity verification module is used for carrying out face verification on the received identity verification information uploaded by the user terminal and carrying out position verification on the identity verification information based on the target image recognition model;
the information acquisition module is used for transmitting the received data information corresponding to the information acquisition request transmitted by the user terminal to the user terminal under the condition that the face verification and the position verification are both passed;
the first training sample is obtained by marking each environment image according to first position information corresponding to each environment image in a plurality of acquired environment images;
the second training sample is determined according to a target image obtained by overlapping the face image and the environment image in the first training sample;
the identity verification information comprises an environment image including a face image of the user and position information of the user, wherein the environment image is shot by the user in real time.
8. An electronic device comprising a processor and a memory storing a computer program, characterized in that the processor implements the information acquisition method of any one of claims 1 to 6 when executing the computer program.
9. A processor-readable storage medium, characterized in that the processor-readable storage medium stores a computer program for causing the processor to execute the information acquisition method according to any one of claims 1 to 6.
10. A computer program product comprising a computer program, characterized in that the computer program, when executed by a processor, implements the information acquisition method according to any one of claims 1 to 6.
CN202210272796.0A 2022-03-18 2022-03-18 Information acquisition methods and systems Pending CN116824314A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210272796.0A CN116824314A (en) 2022-03-18 2022-03-18 Information acquisition methods and systems

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210272796.0A CN116824314A (en) 2022-03-18 2022-03-18 Information acquisition methods and systems

Publications (1)

Publication Number Publication Date
CN116824314A true CN116824314A (en) 2023-09-29

Family

ID=88122672

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210272796.0A Pending CN116824314A (en) 2022-03-18 2022-03-18 Information acquisition methods and systems

Country Status (1)

Country Link
CN (1) CN116824314A (en)

Similar Documents

Publication Publication Date Title
CN113688855B (en) Data processing method, federated learning training method and related devices and equipment
WO2020207189A1 (en) Method and device for identity authentication, storage medium, and computer device
US20230237326A1 (en) Data processing method and apparatus
WO2022105118A1 (en) Image-based health status identification method and apparatus, device and storage medium
TW202026984A (en) Server, client, user verification method and system
CN110322317B (en) Transaction data processing method and device, electronic equipment and medium
CN113435365B (en) Face image migration method and device
CN113591603B (en) Certificate verification method, device, electronic device and storage medium
CN114241459B (en) Driver identity verification method and device, computer equipment and storage medium
CN112650875A (en) House image verification method and device, computer equipment and storage medium
CN112330331A (en) Identity verification method, device and equipment based on face recognition and storage medium
CN111553235B (en) Network training method for protecting privacy, identity recognition method and device
CN114463805B (en) Deep forgery detection method, device, storage medium and computer equipment
CN115905605A (en) Data processing method, data processing equipment and computer readable storage medium
CN116152938B (en) Method, device and equipment for training identity recognition model and transferring electronic resources
CN114861241A (en) Anti-peeping screen method based on intelligent detection and related equipment thereof
CN112765481B (en) Data processing method, device, computer and readable storage medium
CN114359993A (en) Model training method, face recognition device, face recognition equipment, face recognition medium and product
CN116151965B (en) Risk feature extraction method and device, electronic equipment and storage medium
CN115082873B (en) Image recognition method, device and storage medium based on path fusion
CN116824314A (en) Information acquisition methods and systems
CN116681045A (en) Report generation method, report generation device, computer equipment and storage medium
CN115708135A (en) Face recognition model processing method, face recognition method and device
CN114004265A (en) A model training method and node device
CN114120386A (en) Face recognition method, device, equipment and storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination